In analysis of the strengths and weaknesses of the existing workflow model validation methods, a new method is proposed. This is a standardized workflow net based method which can not only detect the regular mistake structures such as deadlock and isolated point but also can detect the AND-split→OR-join(A→O) structure. In order to make detecting A→O structure become reality, by analyzing their features the commonality of various kinds of A→O structures are confirmed and then the appropriate detecting method by detecting the commonality are designed, which means that the method successfully overcome the shortcoming that can not detect the A→O structure of those existing methods. Besides, two normalization rules are built, which can normalize the non-standardized workflow model. It shows that the method is effective and reliable by the case analysis result.

workflow； model normalization； soundness verification； Petri nets